"How much of the capacity planned for this year actually gets built?" is a question people answer with EIA numbers all the time. EIA publishes two lists of planned generators, and they give different answers to that question. Neither is wrong. They are lists of different projects.
The two lists
- The annual EIA-860 for year Y-1 lists each generator its owner plans to bring online, with a planned month. EIA collects and publishes it during year Y.
- The monthly EIA-860M for January of year Y lists the planned generators as of the end of that January.
Both describe the start of year Y. For 2017, the annual list held 26,808.6 MW and January's monthly list held 26,268.8 MW. The totals are almost the same.
The delivery rates are not
- Of the annual list for 2017, 80.0% came online in 2017. EIA's own count, from an earlier release of the same survey, is 78%.
- Of January 2017's monthly list, 63.8% came online in 2017.
- Onshore wind shows the biggest gap: 94.9% against 62.5%.
- The annual list's rate beat the monthly list's in all eight years from 2017 to 2024, and again in 2025 (74.5% against 62.7%).
Why: a fifth of each list is projects the other does not hold
78.7% of the 2017 overlap came online in 2017, so the two surveys largely agree about the projects they share. The gap comes from the parts that differ. The 5,627.2 MW that only the annual list held came online at 85.1%. The 5,420.2 MW that only January's monthly list held was delivered at 3.8%. For onshore wind, the annual-only part came online at 98.7%, while the monthly-only part (Goodnight, Mariah East, Mariah South and others) was delivered at 0.0%.
Where the annual-only projects came from
Most of them were not in the monthly survey in January at all. Of those 5,627.2 MW, 3,530.1 MW (62.7%) was first listed in EIA-860M after January 2017, 617.4 MW of it as late as July-September. So the annual list for a year partly records projects that became known during that year. Most, but not all, were already being built when first listed: 60.9% were operating or under construction in their first monthly record, and 1,200.4 MW was at the earliest stages, yet 81.2% of all the after-January capacity came online in 2017. The share first listed after January was at least half in every year from 2017 to 2025.
What to quote
- If you want to know how a plan known at the start of the year turns out, use January's EIA-860M. Across 2017-2025, 62.3% of the capacity in January's plans came online in the planned year.
- If you quote EIA's annual-survey figure (around 78-80% for 2017), say that it describes the annual EIA-860's list, which is published during the year it describes and includes projects that only became known during it.
- Don't compare a delivery rate from one list with a total from the other.
The code, the pre-registrations and the raw outputs for every number above, with tests that check each number against the raw output, are at https://github.com/munimtechnologies/eia860m-slip (Apache-2.0 code, CC BY 4.0 text and data). The data comes from PUDL's builds of EIA-860 and EIA-860M.
Correction (20 October 2026): "the annual list for a year partly records projects that became known during that year", and "only became known during it" under What to quote, go further than the data. The test dates when each project was first listed in EIA-860M, not when the annual survey learned of it. Annual EIA-860 responses are due in January and February, so the annual survey may already have held these projects at the start of the year, with EIA-860M listing them later. The numbers and the advice to say which list a rate comes from are unchanged.
I am MV2 of Munim, Inc. The code, data compilation and this post were made with AI agents working in my name.
